How do I achieve 90% cost savings with Claude?

By using Prompt Caching on large, static contexts like system instructions or codebases, you can reuse tokens at a significantly lower price point than standard input.

How should I handle Claude API rate limits (429 errors)?

This skill recommends implementing exponential backoff and specifically respecting the 'retry-after' header provided in the API response to maintain stability.

What are Structured Outputs in the Claude API?

Structured Outputs ensure Claude's responses strictly adhere to a provided JSON schema, eliminating parsing errors and providing guaranteed data formats for your application.

Why is my prompt cache not activating?

A common cause is placing the cache_control block incorrectly. It must be placed on the last block of the cacheable content to trigger correctly.

Which models support Structured Outputs?

As of November 2025, Structured Outputs are supported on Claude Sonnet 4.5 and Claude Opus 4 using the structured-outputs-2025-11-13 beta header.

Claude API & Structured Outputs

Name: Claude API & Structured Outputs
Author: jezweb

byjezweb

•

117

•

API Development

Streamlines Claude Messages API integration with structured outputs, prompt caching, and robust error handling patterns.

This skill provides expert guidance for implementing the Anthropic Messages API (v0.69.0+), focusing on high-performance and cost-effective AI development. It enables developers to implement guaranteed JSON schema validation via structured outputs, achieve up to 90% cost savings through optimized prompt caching, and manage complex streaming SSE responses. It is essential for developers building autonomous agents, data extraction pipelines, or chatbots that require strict reliability and efficient context management.

Key Features

01Guaranteed JSON schema validation via Structured Outputs (v0.69.0+)

02Comprehensive migration guides for model deprecations and breaking changes

03Robust error handling for Rate Limits (429) and mid-stream SSE failures

04117 GitHub stars

05Strict Tool Use patterns for reliable function calling and agentic workflows

06Prompt caching implementation for up to 90% input cost reduction

Use Cases

01Building AI agents that require validated JSON data for system integration

02Developing resilient streaming interfaces with proper SSE event parsing

03Reducing operational costs for large-scale applications using context caching

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jezweb/claude-skills claude-api

For use in Claude.ai and ChatGPT

Download Skill